v1.3.3

Top Processes: A remote script now collects the top 5 processes by CPU usage; sensor.<name>_top_processes displays a summary list, while detailed information is provided as attributes including PID, command, CPU, and memory usage.

Adaptive Polling/Backoff: Following repeated errors, the Coordinator temporarily increases the polling interval for each host—up to a maximum of 300 seconds—and resets it once a successful update occurs. The online binary sensor now additionally displays consecutive_failures and current_poll_interval.

Secure run_command Allowlist: The integration options now include a command_allowlist setting. Leaving this list empty ensures backward compatibility, meaning all commands are permitted. Entries must match exactly; entries ending with an asterisk (*) function as a prefix match (e.g., systemctl status *).

Update/Reboot Status Sensors: New sensors have been introduced for each server to track the last package update and the last system reboot; these sensors report a status of success, failed, or never_run, and include success, last_run, and output as attributes.

Full Changelog: v1.3.2...v1.3.3